US 12,314,969 B1
Natural language survey system
Jonathan Robinson, Flushing, NY (US); Leonid Litman, Bronx, NY (US); and Reuben Paris, Flushing, NY (US)
Assigned to Prime Research Solutions LLC, Flushing, NY (US)
Filed by Prime Research Solutions LLC, Flushing, NY (US)
Filed on Nov. 1, 2024, as Appl. No. 18/934,448.
Application 18/934,448 is a continuation in part of application No. 18/766,833, filed on Jul. 9, 2024, granted, now 12,243,066.
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 30/0203 (2023.01); G06F 40/186 (2020.01); G06F 40/58 (2020.01)
CPC G06Q 30/0203 (2013.01) [G06F 40/186 (2020.01); G06F 40/58 (2020.01)] 25 Claims
OG exemplary drawing
 
1. An artificially intelligent system for adaptively generating a customized natural language survey at a large language model, the system comprising:
a processor operating in tandem with the large language model and a database;
a graphical user interface operable to receive a first natural language input, the first natural language input relating to:
one or more survey goals constructed in natural language;
a selection of one or more survey templates;
a set of survey question guardrails constructed in natural language, said set of survey question guardrails comprising:
a description;
an example;
an excluded topic;
an excluded question;
a length limitation;
and an instruction how to manage a specific response by topic;
and/or
an initial question constructed in natural language;
the processor operable to:
translate the first natural language input into a plurality of unstructured survey parameters;
instantiate an instance of an editable version of a survey based on the plurality of unstructured survey parameters, said instance of the editable version of the survey comprising the plurality of unstructured survey parameters;
receive, at the graphical user interface, a second natural language input relating to one or more modifications to the survey;
translate the one or more modifications to one or more unstructured survey parameter modifications;
adaptively modify the instance of the survey by modifying the plurality of unstructured survey parameters with the one or more unstructured survey parameter modifications to generate a modified plurality of survey parameters;
receive, at the graphical user interface, a third input relating to completion of generation of the survey; and
upon receiving the third input, said third input relating to the completion of generation of the survey, convert, at the large language model, the instance of the survey into a conversational survey using the modified plurality of survey parameters by rendering the modified plurality of survey parameters into a series of prompts for use with the large language model;
input the series of prompts into the large language model;
generate, based on output generated from the large language model in response to the input of the series of prompts, a participant instance of the survey for a participant;
assign a memory segment to the participant within the database, said database operating with the large language model;
electronically commence, at the large language model, the participant instance of the survey;
electronically communicate a first survey question to the participant, said first survey question within the participant instance of the survey;
receive a first natural language response from the participant;
auto-format the first natural language response into a data element prompt, said data element prompt constructed in natural language, said data element prompt ingestible by the large language model;
store the first natural language response in the memory segment within the database operating with the large language model, said memory segment assigned to the participant;
adaptively generate, at the large language model, a subsequent survey question, said subsequent survey question based at least in part on information stored within the memory segment assigned to the participant;
receive a subsequent natural language response from the participant;
auto-format the subsequent natural language response into a second data element prompt, said second data element prompt constructed in natural language, said second data element prompt ingestible by the large language model;
store the subsequent natural language response in the memory segment within the database operating with large language model, said memory segment assigned to the participant;
receive, at the large language model, one or more predetermined triggers constructed in structured computer code; and
in response to receipt, at the large language model, of the one or more predetermined triggers constructed in structured computer code, auto-terminate the participant instance of the survey for the participant.